Widefield Quantum Sensor for Vector Magnetic Field Imaging of Micromagnetic Structures

TL;DR

This work tackles the challenge of wide-area, vector-resolved magnetic-field imaging at micrometer scales by implementing a camera-based pulsed ODMR protocol on a commercial widefield microscope with a shallow NV layer. By measuring the Zeeman shifts along the four NV ⟨111⟩ orientations, the authors reconstruct the full 3D magnetic-field vector $\mathbf{B}$ across the field of view, demonstrated on microfabricated permalloy ellipses. They achieve a spatial resolution of about $0.52~\mu\mathrm{m}$ over an $83\,\mu\mathrm{m} \times 83\,\mu\mathrm{m}$ field of view with per-orientation sensitivities of roughly $0.8$–$2.1~\mu\mathrm{T}/\sqrt{\mathrm{Hz}}$, acquiring complete vector maps in minutes. The approach offers a practical, scalable platform for rapid vector-resolved imaging of complex magnetic devices on standard optical microscopes, with potential applications in skyrmions, 2D magnets, and neuromorphic/spintronic systems.

Abstract

Many spintronic, magnetic-memory, and neuromorphic devices rely on spatially varying magnetic fields. Quantitatively imaging these fields with full vector information over extended areas remains a major challenge. Existing probes either offer nanoscale resolution at the cost of slow scanning, or widefield imaging with limited vector sensitivity or material constraints. Quantum sensing with nitrogen-vacancy (NV) centers in diamond promises to bridge this gap, but a practical camera-based vector magnetometry implementation on relevant microstructures has not been demonstrated. Here we adapt a commercial widefield microscope to implement a camera-compatible pulsed optically detected magnetic resonance protocol to reconstruct stray-field vectors from microscale devices. By resolving the Zeeman shifts of the four NV orientations, we reconstruct the stray-field vector generated by microfabricated permalloy structures that host multiple stable remanent states. Our implementation achieves a spatial resolution of $\approx 0.52 ~μ\mathrm{m}$ across an $83~μ\mathrm{m} \times 83~μ\mathrm{m}$ field of view and a peak sensitivity of $ (828 \pm 142)~\mathrm{nT\,Hz^{-1}}$, with acquisition times of only a few minutes. These results establish pulsed widefield NV magnetometry on standard microscopes as a practical and scalable tool for routine vector-resolved imaging of complex magnetic devices.

Figures (5)

• Figure 1: Illustration of nitrogen-vacancy (NV) vector magnetometry for imaging magnetic structures. A near-surface NV layer in diamond senses the stray field from microstructures, with each NV orientation measuring a different projection and enabling reconstruction of the magnetic-field vector.
• Figure 2: Experimental setup for widefield NV-based magnetometry. (a) Energy level structure of the NV center with a triplet ground and excited state, and two intermediate states. Relevant transitions are highlighted by arrows. Inset: splitting of the magnetic spin states $m_\mathrm{s}=\pm1$ due to the Zeeman effect. (b) Optical microscope image of two crossing ellipse structures (left) and an Object Oriented MicroMagnetic Framework simulation of one of the four possible remanent states (right). (c) Schematic of the experimental NV-based setup. Inset: cross-section showing the sample configuration with magnetic structures placed above the diamond and microwave antenna. (d) Pulsed ODMR protocol for widefield imaging. Each sequence is repeated $\mathrm{n}_\mathrm{r}$ times to match the exposure time of the camera (top). Sequence of signal acquisition alternating between Signal and Reference protocols, with the microwave frequency sweep from point 1 to $m$, repeated $N$ times (bottom).
• Figure 3: Experimental Results from NV-Based Magnetometry. (a) Fluorescence image of the two crossing ellipses with two 10 × 10 pixels regions highlighted, blue and red. (b) Protocol implemented to measure the Rabi oscillations (top) and the respective oscillations obtained for the $[1\bar{1}\bar{1}]$ orientation (bottom). The $\pi$ pulse is represented in the figure as the first minimum point of fluorescence corresponding to 110 ns. (c) ODMR spectra for the two regions highlighted in (a). A shift in the resonance frequency is visible between the two areas, influenced by the magnetic field produced by the ellipses.
• Figure 4: Frequency maps obtained for the four NV orientations ([111],$[1\bar{1}\bar{1}]$, $[\bar{1}1\bar{1}]$, $[\bar{1}\bar{1}1]$). The frequency variation is confined to the regions inside the ellipses in all cases.
• Figure 5: Maps of the magnetic field vector components: (a) x-component, (b) y-component, and (c) z-component of the magnetic field reconstructed from the four NV orientations. A binning of 3x3 pixels was applied to retain spatial resolution in the reconstructed maps. (d) Map of the magnetic field strength obtained by combining the three vector components. The dashed gray lines delimit the magnetic structure. The inset displays the in-plane field direction in the central region, derived from the $\mathrm{B}_\mathrm{x}$ and $\mathrm{B}_\mathrm{y}$ components.